Ronald Stenvert
When working with the computer in relation to art historical subjects, and being interested in the content of the image itself, one is caught between rudimentary image-processing on the one hand. and imprecise translation into words on the other hand. It will never be completely possible to bridge the gap between the resulting extremes of global descriptions and technical algorithms. In this contribution some theoretical reflections upon the structuring of art historical data before input will be discussed, followed by the outlines of some 'building blocks' to bridge the gap between pixel and picture, and its possible effects on the shifting scope of research in art history.